Abstract: Apparatus and method are disclosed for achieving automated evaluation and oring of a participant's response to questions selected pseudo-randomly. A plurality of questions are available for each of a plurality of categories, in a question file. For each question there is an answer in an answer file. The answer includes characters preselected to denote a tolerance that permits a partially correct answer automatically to be credited to the participant's point total. A perfect answer is worth a predetermined full-credit value that is the same for each question; and, a partially correct answer that is within the tolerance for that question, is worth a predetermined fraction of the full-credit value. Pseudo-random selection of the questions from the question file is performed by a number generator that is seeded from the system's clock with the time of day that the system is activated.